GemClass

GemClass is an AI-powered tutoring system that uses Google's Generative AI to provide personalized educational assistance. The system is designed to guide students through their learning process without simply giving away answers, encouraging critical thinking and deep understanding. Made by Sebastian Castro (sec204@pitt.edu) and Daley Fraser (drf58@pitt.edu)

Table of Contents

1. Project Overview
2. Features
3. Technology Stack
4. Project Structure
5. Setup Instructions
6. Usage
7. Customization
8. Planned Updates

Project Overview

GemClass is built to simulate a knowledgeable tutor that adapts to different courses and helps students learn effectively. It uses a chat-based interface where students can ask questions, and the AI responds with guidance, hints, and explanations tailored to the selected course.

Features

• Course-specific AI tutoring
• General AI tutoring
• Chat-based interface for easy interaction
• Markdown support for formatted responses

Technology Stack

• Backend: Python with Flask
• Frontend: HTML, CSS, JavaScript
• AI: Google's Generative AI (Gemini model)
• Markdown Parsing: Marked.js library

Project Structure

gemclass/
│
├── app.py                 # Main Flask application
├── gemini_test.py         # Gemini API integration
├── frontend/
│   └── index.html         # Main HTML template
├── static/
│   └── steelhacker.png    # Logo image
└── class_data/
    ├── cs50.json          # Harvard CS50 course data
    └── cs447.json         # Pitt CS0447 course data

Built With

• flask
• html
• python